52 I The Process of Evolution 



CONTINUOUS VARIATION 



When one comes to study the genetics of continuously varying char- 

 acteristics, the problems become much more difficult. Usually it is 

 not possible to identify specific genes controlling specific traits. As 

 has been mentioned above, there is thought to be a continuous spec- 

 trum of characters ranging from those which vary qualitatively to 

 those which vary quantitatively. Presumably there is no basic differ- 

 ence between genes with easily detected major eflFect (often called 

 oligogenes or switch genes) and those with only minor efiFect indi- 

 vidually but which operate as part of a system of an indefinite num- 

 ber of factors (often called polygenes, although this term has been 

 used in a more restricted sense). Operationally, it may be said that 

 the difference between the two types of characters depends more on 

 the relative importance of the genetic material and the environment 

 in determining the phenotype than upon the size of individual gene 

 effects. However, there are polygenic characters in which the role 

 of the environment is relatively unimportant, such as number of 

 abdominal bristles in Drosophila. 



In studying variation in a quantitative character, as large a sample 

 as possible of differing individuals is measured. All the individuals 

 are unique but their measurements may be grouped into size classes. 

 When these measurements are plotted as a frequency distribution, 

 the nature of the variation may be studied. For instance, the arith- 

 metic mean or average may be calculated for all the individuals in 

 the sample. The amount of variation in the sample may be estimated 

 by the standard deviation S or its square, the variance S-. 



It is usually difficult to separate genetic and environmental com- 

 ponents of variation and to study the genetic portion independently. 

 Various laboratory techniques and rather complex mathematical 

 formulations have been developed to study and separate these com- 

 ponents. For example, one way to estimate the size of these two 

 components involves the reduction of the genetic component until 

 it is negligible. The variance in the character measured is observed 

 in a population in a "normal" environment. Then the variance of the 

 same character is measured in individuals of a highly inbred line 

 raised in the same environment. Since these individuals may be con- 

 sidered to be essentially identical genetically, the variance observed 

 may be attributed entirely to the effects of environment. The differ- 

 ence between the two variances is then an estimate of the genetic 

 variance, since total variance (phenotypic variance S^) is, in this 



